An integrated translational research program has been developed with the goal of designing, developing, and delivering new immunotherapies for cancer treatment and prevention with emphasis on immune intervention for a range of human carcinomas. Vector-based delivery of tumor-associated antigens, T-cell costimulatory molecules, and cytokines in the induction of immune responses and anti-tumor immunity-experimental studies: Tumor-associated antigens (TAAs) are by definition either weakly immunogenic or functionally non-immunogenic in the immune competent host. Thus, if cancer vaccines are to be effective in either preventing or eliminating tumors, vaccines and vaccine strategies must be developed to present antigens to the immune system in a way that makes them more immunogenic. Efforts have focused on the development of strategies that convert poorly immunogenic antigens into more potent immunogens in the induction of host T-cell responses. Studies have now shown that two classes of recombinant poxvirus vectors (vaccinia and avipox) that contain transgenes for tumor antigens can render those gene products more immunogenic. Recombinant vaccinia vectors have been shown to be potent in the induction of the immune response, while recombinant avipox vectors, which are replication defective, have been shown to be ideal for multiple boosting in the absence of host-limiting immunity to the vector. The need for T-cell costimulation has been shown to be extremely important in the generation of potent T-cell responses, especially when the antigen in question is weakly immunogenic. Recombinant poxvirus vectors have been constructed that contain one, two or three transgenes for T-cell costimulatory molecules. Recombinant vectors containing a triad of costimulatory molecules (B7-1, ICAM-1, and LFA-3, designated TRICOM) have been shown to be extremely effective in enhancing host T-cell responses. Peptide-pulsed dendritic cells or B cells that have been infected with avipox-TRICOM vectors have been shown to be far more effective in the induction of both naive and memory T-cell responses (both in vitro and in vivo) than the use of control peptide-pulsed dendritic cells or B cells, respectively. The mechanism by which T-cell costimulation enhances T-cell activation has also been studied. It has now been demonstrated that T cells actually acquire T-cell costimulatory molecules from antigen-presenting cells upon activation. Depending on the level of signal 1 and the level of expression of costimulatory molecules, this phenomenon can have either immune stimulatory or immune regulatory consequences. Recombinant avipox vectors have now been constructed which contain the cytokine GM-CSF. Studies have now demonstrated that the use of this vector along with recombinant vaccines greatly enhances the number of dendritic cells in nodes proximal to the vaccination site, with consequent increases in both host immune responses and anti-tumor activity. Development of more valid animal models for the analysis of new vaccine strategies: The availability of valid animal models that mirror the human cancer patient is extremely important in the analysis of new vaccines and vaccine strategies. Carcinoembryonic antigen (CEA) is overexpressed on a wide range of human carcinomas, including colorectal, pancreatic, breast, non-small cell lung, and head and neck tumors, and to a lesser extent on normal colonic epithelium. A CEA transgenic mouse, which expresses CEA in a manner similar to that found in humans, has been utilized for the analysis of induction of immunity and anti-tumor effects using recombinant CEA-based vaccines. These studies have now demonstrated that poxvirus-based CEA/TRICOM vectors can eliminate established tumors that could not be eliminated with less potent vaccines and vaccine strategies. Recent studies have also demonstrated that TRICOM-based vaccines can also prevent the induction of spontaneous tumors in other mouse models. Design and development of recombinant immunoglobulin molecules: Efforts have focused on the development of novel recombinant immunoglobulin forms that can be used for the detection and/or therapy of human carcinomas. MAbCC49, when used as a radioimmunoconjugate, has previously been shown to efficiently target human ovarian, prostate and colorectal cancers in clinical trials. A CDR-grafted "humanized" version of CC49, devoid of the CH2 domain, has now been developed (HuCC49DCH2). This molecule has the property of rapid clearance from the body, and reduced immunogenicity, properties that make it more suitable as a radioimmunotherapeutic. Recombinant variants of this molecule have also been developed that have enhanced affinity for the target antigen. Activation of human T cells by carcinoma-associated antigens: The primary focus of these studies is to identify those epitopes of tumor-associated antigens expressed on human carcinomas that have the ability to activate human T cells. Several different epitopes of human CEA have been identified that have the ability to activate human T cells, which in turn can lyse human carcinoma cells. An agonist epitope of one of these determinants has now been identified that has the ability to activate T cells to far greater levels than the native epitope, and has the ability to kill tumor cells expressing native CEA. Similar studies have also been conducted with human prostate-specific antigen (PSA) and the MUC-1 antigen expressed on a wide range of human carcinomas. Human TRICOM vectors have now been developed that simultaneously express three human costimulatory molecule transgenes. Peptide-pulsed human dendritic cells, infected with TRICOM vectors, are far more effective in the activation of human T cells than the use of peptide-pulsed dendritic cells. These findings along with those of preclinical murine models using TRICOM vectors have led to the translation of these studies to ongoing clinical trials. Development of vaccines for clinical trials/clinical trials program: As a result of hypothesis-driven preclinical studies, a series of recombinant vaccines has been designed and developed for use in clinical trials. These are both peptide-based and vector-based vaccines, which are directed against three different carcinoma-associated antigens: CEA, PSA, and MUC-1. Vector-based vaccines not only contain transgenes for tumor-associated antigens, but also transgenes for multiple costimulatory molecules (i.e., TRICOM vectors) as well as cytokine genes. A series of immunoassays has now been developed to analyze patients' immune responses both prior to and post-vaccination. Collaborative vaccine clinical trials are currently ongoing in collaboration with the NCI CCR Medical Oncology Research Unit and eight other Cancer Centers throughout the country. These trials have demonstrated the ability of advanced carcinoma patients to mount T-cell responses to "self" tumor-associated antigens as a consequence of vaccination. Phase I/II clinical trials have now demonstrated the advantage of (a) the use of recombinant poxvirus-based vaccines, (b) diversified prime and boost vaccine regimens, and (c) the use of agonist epitopes, in enhancing the generation of T-cell responses and in achieving objective clinical responses and/or increased survival in advanced carcinoma patients.